This research program grant application (R01) requests funds to conduct basic investigation of how social networks affect risk and risk reduction practices longitudinally in the context of sex-drug use within a community-based sample of younger Black men (16-29) who have sex with men (YBMSM) in order to identify critical points for intervention to reduce HIV transmission. YBMSM are embedded within heterogeneous Black communities. From the perspective of an individual YBMSM, his social network often include close confidants such as primary partners, relatives and friends who may exert supportive or antagonistic influences on his behavior and choice of partner(s). Many of these friends may crossover between the non-sexual and sexual components of one's network over time. Additionally, numerous other structures and institutions such as the ballroom community, detention centers/jails and internet social networking sites shape YBMSM networks and norms. How YBMSM are positioned within these rich and dynamic social environments and the considerable influence their social networks have on sex partner selection, sexual/drug related practices and risk-reduction behavior with specific partners has only recently begun to be explored. Our analytic approach focuses on components of the Auerbach 'social drivers' model applied to YBMSM, with particular attention to two components that this team is well positioned to study: norms and networks. These factors shape individuals' behaviors and practices modulated by situational factors such as sex-drug use. Based upon our preliminary work, we found that 1) dyadic mixing by risk behaviors (UAI, sex-drug use, group sex); 2) influences from social network members; and 3) composition of non-sexual social networks are associated with risk and biomedical prevention interventions in cross-sectional analyses of YBMSM. Thus far however, causal associations between networks and other 'social drivers' of HIV infection among BMSM, have been elusive. Accordingly, utilizing a 3 wave longitudinal cohort approach we aim to: 1) Describe the dynamics of social networks among YBMSM ages 16-29, and estimate the relationships between specific network attributes (e.g., confidant composition and component size) and subsequent risk/risk reduction behaviors, including their interaction with various social structures and institutions; 2) Examine the relationships between social networks, norms, risk/risk reduction practices and HIV/STI infection over time among this sample and their extended networks; 3) Determine the extent to which situational factors such as sex-drug use mediates the relationship between social network characteristics, norms and risk/risk reduction practices over time; and 4) Explore how learning that one is HIV positive affects social networks, norms and risk/risk reduction practices. This research is a necessary step in the development of scientifically-based network interventions that can complement combination interventions targeted to diverse YBMSM groups and their extended networks.